RESEARCH / VERSION 1.1 / 16 OCTOBER 2026
Recovery at work
What support do employers put in writing?
Coverage: 30 reviewed public employer documents. A purposive pilot, not a representative survey or a guarantee of current HR practice.
Findings and their limits
What 30 public employer documents put in writing about treatment time and recovery. The sample contains 13 universities, ten NHS bodies, five councils and two other bodies; it is not representative of UK employers.
27 documents describe a treatment-time route. Seven contain express paid treatment/appointment wording; six further records expressly use sick-pay terms. Fourteen identify a route without establishing payment, and three lack an identified treatment-time provision within the examined scope. These categories describe wording, not staff experience or unconditional entitlements.
| Provision | Explicitly identified | Not identified in scope | Not reviewed |
|---|---|---|---|
| Confidentiality | 30 | 0 | 0 |
| Voluntary help-seeking | 29 | 1 | 0 |
| Treatment time | 27 | 3 | 0 |
| Relapse / recurrence | 21 | 8 | 1 |
| Return / rehabilitation | 27 | 3 | 0 |
| Safety / impairment | 29 | 1 | 0 |
| Occupational health | 28 | 2 | 0 |
Conditions are recorded separately. “Explicit” includes conditional provisions; “not identified” is not “never offered”. The seven express-paid passages were rechecked on 16 October; the sick-pay category is a recode of existing evidence descriptions. Other leave terms may exist outside reviewed documents.
New checks identified Hywel Dda V6 and its treatment guidance, resolved Public Health Wales’s lapse diagram and examined Salisbury’s three appendices. Policy currency and implementation remain separate questions.
Cite and reuse
Civil Liberties Action (2026). Recovery at work [Dataset]. Version 1.1, 2026-10-16. https://cla-political-finance.copperchime4.chatgpt.site/databases/releases/2026-10-16-v1.1/workplace-research.html
BibTeX RIS Full CSV Full JSON Reproducibility package
For an individual observation, cite its fixed-version record and the original source. Credit original publishers for their figures; CLA supplies the compilation, coding and stated analysis. Reuse terms.
Methods and selection
Unit of analysis
One employer and the identified policy version, with separately identified supporting documents. Seven provision dimensions; a missing statement is not evidence of absence in practice.
Inclusion and coverage
30 accessible public-sector/publicly funded employer documents: 13 universities, 10 NHS bodies, five councils and two other bodies. Purposive discovery sample; not a census or representative survey. Employer-hosted substantive text was required. Inaccessible documents, unfilled templates and unconfirmed drafts were excluded rather than coded as no policy.
Coding and checks
- Legacy status is preserved for traceability. Presence is explicit for both legacy yes and conditional; otherwise not stated or not reviewed. Conditions are recorded separately. An explicit provision is not necessarily an unconditional entitlement.
- Treatment leave types distinguish ordinary sickness/attendance, annual leave, paid appointments, unpaid leave and other discretionary time. Only stated arrangements are coded; not established remains available. Pay is separate from type. These are descriptions of documents, not enforceability ratings.
- An old or overdue review date does not prove invalidity. The default currency status is published document examined; current internal operation is unconfirmed unless direct evidence establishes it.
- The 16 October recheck resolved supporting-document gaps for Public Health Wales and Salisbury and identified Hywel Dda V6 and its operational guidance. The change log identifies affected codes. Other source dates have not been refreshed automatically.
Selection-history limitation
The first pilots were assembled through purposive discovery and source checking. Their complete original search-engine query/results history was not retained, so the first selection cannot be replayed exactly. This limitation is not repaired by writing a retrospective systematic-search claim. Published records and transformations can be reproduced from the supplied files; prospective expansion protocols below define the next collection work.
Prospective expansion protocol
Before expanding, select a complete named employer population and record its membership and cutoff date. Search each employer's official website for alcohol, drugs, substance misuse and recovery policies, follow appendices, and log documents found, blocked, excluded and unresolved. Have a second human coder independently assess a sample before describing inter-rater reliability; repeated AI checks are not independent human review.
Interpretation limits
Written policies cannot establish implementation, staff experience, effectiveness, legal compliance or comparative employer quality. Northern Ireland policy records do not import Great Britain equality legislation.
Dates and accountability
Release date is the date of this publication. reviewed_at records a stated evidence check, not proof that every fact remains current. Publication date, event date, reference period, document version and source-check date are separate fields.
Published by Civil Liberties Action. Compilation, source retrieval and drafting used AI assistance. Automated integrity checks and additional source checks are described below; no independent external or specialist review is claimed.
Download the protocol and dictionary
Data dictionary
| Field | Definition |
|---|---|
| id / record_id | Stable record identifier; retained across releases where the unit remains the same. |
| release / version | Publication date and release version, separate from the date of underlying evidence. |
| record_url / citation | Fixed-version record link and suggested citation. |
| category / geography / period | Display group, actual coverage and source reference period; not a harmonised statistical panel. |
| reviewed_at | Date of the recorded source check, subject to its stated scope. |
| details / sources | Labelled observations and original URLs, titles and locators. |
| limitations | Record-specific boundary on interpretation. |
| raw | Underlying typed source/coding fields; null remains missing, not zero. |
| provision_coding.presence | explicit, not_identified_in_reviewed_scope or not_reviewed. |
| condition_status | condition_or_scope_recorded; not_separately_assessed; not_applicable; unresolved. Not a quality score. |
| treatment_leave_classification | Multiple leave routes; separate pay_status, pay_note, source locator and any targeted check date. |
| provisions.status | Legacy coding retained; yes and conditional both indicate explicit presence. |
Fixed-version records
| Record | Reference period | Status |
|---|---|---|
| University of Dundee | 2023-06-01 | Published document examined; internal operation unconfirmed |
| University of Exeter | Document date not established | Published document examined; internal operation unconfirmed |
| Falkirk Council | Document date not established | Published document examined; internal operation unconfirmed |
| Rotherham Doncaster and South Humber NHS Foundation Trust | 2024-02-29 (version 6) | Published document examined; internal operation unconfirmed |
| Leeds Community Healthcare NHS Trust | 2026-01-26 | Published document examined; internal operation unconfirmed |
| Leeds Beckett University | 2019-12 (V1219) | Published document examined; internal operation unconfirmed |
| Keele University | 2023-12-19 (version 1.3) | Published document examined; internal operation unconfirmed |
| Loughborough University | Document date not established | Published document examined; internal operation unconfirmed |
| Oxford Brookes University | 2010-11 | Published document examined; internal operation unconfirmed |
| University of Bristol | Document date not established | Published document examined; internal operation unconfirmed |
| Louth Town Council | 2024-04-30 | Published document examined; internal operation unconfirmed |
| West Lothian Council | 2019-04-23 | Published document examined; internal operation unconfirmed |
| Sharlston Parish Council | Document date not established | Published document examined; internal operation unconfirmed |
| Lincolnshire County Council | Document date not established | Published document examined; internal operation unconfirmed |
| Tees, Esk and Wear Valleys NHS Foundation Trust | 2025-07-15 (version 1) | Published document examined; internal operation unconfirmed |
| Surrey and Borders Partnership NHS Foundation Trust | 2026-10-07 (version 6.0; approved 2026-08-06) | Published document examined; internal operation unconfirmed |
| Royal Cornwall Hospitals NHS Trust | 2025-02 (version 4.0) | Published document examined; internal operation unconfirmed |
| University of Bath | 2024-05-01 | Published document examined; internal operation unconfirmed |
| University of St Andrews | 2025-06-02 (version 1.5) | Published document examined; internal operation unconfirmed |
| University of Glasgow | 2023-08 | Published document examined; internal operation unconfirmed |
| Velindre University NHS Trust | 2024-09-26 (version 4, WF18) | Published document examined; internal operation unconfirmed |
| Public Health Wales NHS Trust | 2022-09-06 approved; published 2022-09-12 (version 3, PHW49) | Published document examined; internal operation unconfirmed |
| Hywel Dda University Health Board | 2026-05-19 approved; active 2026-09-11 (version 6, policy 283) | Published document examined; internal operation unconfirmed |
| Aberystwyth University | 2019-07 last reviewed (version 1.1) | Published document examined; internal operation unconfirmed |
| Ulster University | 2014-06-03 last edit; effective 2014-08-01 (version 3.1) | Published document examined; internal operation unconfirmed |
| Probation Board for Northern Ireland | 2019-02-22 board approval; implemented 2019-03-01 (version 0.3) | Published document examined; internal operation unconfirmed |
| Sport Wales | Document date not established | Published document examined; internal operation unconfirmed |
| Cardiff and Vale University Health Board | 2023-01-18 approved; published 2023-01-31 (version 4, UHB171) | Published document examined; internal operation unconfirmed |
| Queen Margaret University | 2017-01 last reviewed; issued 2012-11-30 | Published document examined; internal operation unconfirmed |
| Salisbury NHS Foundation Trust | 2021-05 (version 1.2) | Published document examined; internal operation unconfirmed |
Corrections, funding and interests
Send the record identifier, proposed correction and supporting primary source to policy@civillibertiesaction.org.uk. Corrections are assessed against the same evidence rules for all subjects. Responses, reversals and corrected outcomes receive the same factual prominence as initial decisions. Substantive changes require a dated new release and a change-log entry; the old files remain available.
CLA's existing funding and interests statement applies. Its founder has disclosed Labour Party membership. The same inclusion, identity and correction rules apply regardless of political affiliation. Read the declaration.